[
https://issues.apache.org/jira/browse/FLINK-39122?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18088952#comment-18088952
]
Rion Williams commented on FLINK-39122:
---------------------------------------
[~twalthr]
I had some time today to revisit this and add some appropriate detection for
invalid decimals during type-inference. I've crafted a proposed PR for it along
with a suite of tests to verify it and tagged you on it as well:
* [Pull Request|https://github.com/apache/flink/pull/28437]
Any/all feedback is welcome and greatly appreciated!
> Decimals with incorrect precision lead to null value in PTF
> -----------------------------------------------------------
>
> Key: FLINK-39122
> URL: https://issues.apache.org/jira/browse/FLINK-39122
> Project: Flink
> Issue Type: Sub-task
> Components: Table SQL / Planner
> Reporter: Timo Walther
> Priority: Major
> Labels: pull-request-available
>
> Instead of a type error the following call leads to a null value for the
> decimal:
> {code}
> // execute a Flink SQL job and print the result locally
> tableEnv.fromCall(RulePtf.class, new
> BigDecimal("123.456")).execute().print();
> }
> /** A PTF that filters data based on a dynamic allowlist. */
> public static class RulePtf extends ProcessTableFunction<String> {
> public void eval(@DataTypeHint("DECIMAL(2, 2)") BigDecimal decimal)
> throws Exception {
> System.out.println(decimal);
> }
> }
> {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
